Only do rolling restart can't solve problem. But I thought I found a way.
There are two same name folder with different suffix in data dictionary. e.g.
dayu_123 and dayu_234, the dayu_123 folder is empty and dayu_234 folder is not.
then I use cql to query system_schema.tables，the id of table
Yeah, but you only really need to drain, restart Cassandra one by one. Not
that the others will hurt, but they aren't strictly necessary.
On 28 June 2018 at 05:38, dayu wrote:
> Hi kurt, a rolling restart means run disablebinary, disablethrift,
> disablegossip, drain,
> stop cassandra and
Best off trying a rolling restart.
On 28 June 2018 at 03:18, dayu wrote:
> the output of nodetool describecluster
> Cluster Information:
> Name: online-xxx
> Snitch: org.apache.cassandra.locator.DynamicEndpointSnitch
> Partitioner: org.apache.cassandra.dht.Murmur3Partitioner
> Schema versions:
